India's verification landscape is being transformed by multi-agent AI systems that analyse careers holistically rather than checking documents one by one.
For most of the past two decades, employment verification in India followed a straightforward script. You collected a candidate's documents, called the HR number listed on the company website, waited for a callback, and noted whatever the HR person said. The entire process was linear, manual, and deeply dependent on whether someone picked up the phone. In 2026, AI has fundamentally changed every part of this process. AscentPassport uses six specialist AI agents that run in parallel on every candidate profile. Each agent is responsible for a specific dimension of verification, and their outputs are combined by a weighted orchestration system into a single AP Score. Understanding what each agent does helps explain why this approach catches fraud that traditional verification misses entirely. The HR Confirmation Agent manages automated outreach to previous employers. Rather than relying on a human recruiter to find the right HR contact and make a call that may go unanswered, this agent drafts a professional verification email, sends it to the confirmed HR contact at the previous employer, and tracks the response. When HR confirms, the experience is marked verified and the AP Score receives its largest single boost. When they report a discrepancy, the score reflects it immediately. The whole cycle runs in hours rather than weeks. The Timeline Consistency Agent analyses the full employment history as a connected narrative rather than isolated records. It checks for date overlaps where two jobs appear to run simultaneously, gaps longer than six months that are unaccounted for, title regressions that suggest fabricated progression, and tenure durations that are implausible for the claimed role. A candidate with ten years of experience who claims to have held seven Director-level positions at seven different companies triggers an immediate consistency review. The Government ID Agent validates the candidate's PAN or passport number and anchors the entire profile to a single verified real-world identity. This prevents duplicate accounts, stops identity sharing, and ensures that an APID cannot be used by anyone other than the verified individual. The format validation used goes beyond simple regex checks to include checksum verification specific to Indian PAN number conventions. The Digital Footprint Agent cross-checks the professional's online presence against their claimed employment history. LinkedIn profiles, GitHub repositories, company websites, and MCA registry data are all assessed for consistency with the candidate's stated career. A candidate who claims five years at a company that appears nowhere in any public database, has no LinkedIn presence, and has no MCA registration is flagged for further review. The Financial Signals Agent assesses salary ranges and provident fund data for plausibility. Salary claims are evaluated against role, company size, location, and tenure. PF account numbers are validated for format consistency with EPFO conventions. Significant discrepancies between claimed salary and what the role and market data would support generate a flag. The Profile Completeness Agent rewards thorough, honest profiles. A professional who provides detailed role descriptions, accurate HR contact details, PF numbers, and supporting information receives higher scores because statistical analysis shows that comprehensive profiles correlate strongly with accurate ones. The combined output of these six agents produces the AP Score and a detailed breakdown that any employer can inspect. This level of systematic, multi-dimensional verification is something no traditional BGV report has approached. For hiring managers who want a practical sense of what this means: checking a candidate's APID takes under two minutes. The AP Score and dimension breakdown are visible immediately. If the score is above 75 with strong HR confirmation, you have more verified information about that candidate's history than any traditional BGV report would have provided, and you have it before the first interview. The verification work having been done in advance fundamentally changes what the interview itself is for. The six-agent architecture is designed to be modular. As new verification signals become available, additional agents can be added without disrupting the existing scoring framework. This forward-compatible design means the AP Score becomes more reliable over time as the platform adds capability, not less reliable as fraud methods evolve.